Deep Learning in Cardiovascular Disease Prediction: A Survey of Literature

Abstract

One essential organ crucial to preserving general human health is the heart. Modern lifestyle and a host of associated factors have led to an increase in the number of people having cardiovascular disease. Nowadays, in comparison to Cancer-related deaths, more people die from cardiovascular diseases (CVDs), thereby making it a major global health concern. To lower morbidity and mortality, early detection and precise prediction of CVDs are essential. In this regard, medical professionals can identify a variety of cardiac conditions, such as heart failure and valve disorders, with the help of computer-aided diagnostic systems. The enormous amount of data generated every day provides ample scope to utilize tools such as data mining to gather valuable insights in this field, particularly when we are well-equipped for big data analysis. Further, examining a variety of risk factors, Deep learning techniques can help us to use a large number of parameters to predict various kinds of heart diseases well in advance. The present paper seeks to evaluate and compare the predictive performance of different deep learning models for the early diagnosis of cardiovascular conditions. These models include convolutional neural networks (CNNs), Intermittent neural networks (RNNs), Transformers, Generative Adversarial Networks (GANs), and Auto Encoders. The paper would highlight how efficiently these models predict heart disease. This would further open the scope for the creation of ‘sophisticated clinical decision support system’ that focuses on early detection and prevention of cardiovascular disease.